SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.01081 · 2024

LAB: Large-Scale Alignment for ChatBots

arXiv · PDF · Open in the Atlas

Code that ran

4 of 5 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 4 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2403.01081.json · badge: /api/ran/2403.01081.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
instructlab/instructlab community (classifier label)
classifier basis not recorded on the link
4 ran · 1 no recorded run
FunctionStatusWhere it lives
display_params Ran
ran on a synthesized input
instructlab/instructlab/src/instructlab/clickext.py
code served (permissive licence) · get_code("d2a87e42b1fc64de")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned function. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import functools
import json
import logging
import os
import sys
import typing
import click

# Set up logging
logging.basicConfig(level=logging.DEBUG)
logger = logging.getLogger(__name__)

# Create a simple test function to decorate
@click.command()
@click.option('--name', default='test', help='A name parameter')
@click.option('--count', default=1, type=int, help='A count parameter')
@click.pass_context
def test_function(ctx, name, count):
    """Test function for display_params decorator"""
    return f"name={name}, count={count}"

# Build the decorator
from functools import wraps

def _get_param_info(ctx: click.Context, **kwargs: dict) -> typing.Generator[tuple[str, typing.Any, str, str], None, None]:
    """Get click parameter information"""
    for key, value in kwargs.items():
        src = ctx.get_parameter_source(key)
        param_src: str = src.name.lower() if src is not None else 'unknown'
        type_name: str
        if value is None:
            type_name = 'None'
        else:
            type_: type = type(value)
            type_name = type_.__name__
            mod_name: str | None = getattr(type_, '__module__', None)
            if mod_name and mod_name != 'builtins':
                type_name = f'{mod_name}.{type_name}'
        yield (key, value, type_name, param_src)
    if ctx.args:
        yield ('args', ctx.args, 'list', 'args')

class _ParamEncoder(json.JSONEncoder):
    """Custom encoder for additional parameter types"""
    def default(self, o: typing.Any) -> typing.Any:
        if isinstance(o, os.PathLike):
            return os.fsdecode(o)
        return super().default(o)

def display_params(f: typing.Callable) -> typing.Callable:
    """Display command parameters decorator"""
    @functools.wraps(f)
    def wrapper(*args: typing.Any, **kwargs: typing.Any) -> typing.Any:
        ctx: click.Context | None = kwargs.get('ctx')
        if ctx is None:
            ctx = click.get_current_context()
        display: str = kwargs.pop('_debug_params', None)
        if display is None and logger.isEnabledFor(logging.DEBUG) and (not kwargs.get('quiet', False)):
            display = 'logger'
        if kwargs and display in {'logger', 'human'}:
            pad = min(max((len(name) for name in kwargs)), 24)
            print('Parameters:')
            for name, value, typ, src in _get_param_info(ctx, **kwargs):
                print(f'  {name:>{pad}}: {value!r:<8} \t[type: {typ}, src: {src}]')
        elif display == 'json':
            params = {name: (value, type_name, param_src) for name, value, type_name, param_src in _get_param_info(ctx, **kwargs)}
            json.dump({'params': params}, sys.stdout, cls=_ParamEncoder)
        if display in {'json', 'human'}:
            ctx.exit()
        return f(*args, **kwargs)
    human_option = click.option('--debug-params', '_debug_params', flag_value='human', hidden=True, help='display parameter info and exit.')
    json_option = click.option('--debug-params-json', '_debug_params', flag_value='json', hidden=True, help='display parameter as JSON and exit.')
    return human_option(json_option(wrapper))

# Create a mock click context
mock_ctx = click.Context(click.Command('test'))

# Arguments for display_params function
ARGS = {
    'f': test_function,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
click — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
expand_path Ran
ran on a synthesized input
instructlab/instructlab/src/instructlab/utils.py
code served (permissive licence) · get_code("323d0c3f78f00387")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os

# Set up environment variable for testing
os.environ['TEST_VAR'] = 'test_value'

# Test with a simple path string containing ~ and environment variables
ARGS = {
    'path': '~/documents/$TEST_VAR/file.txt'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_ssl_cert_config Ran
ran on a synthesized input
instructlab/instructlab/src/instructlab/client_utils.py
code served (permissive licence) · get_code("d31a01a2351d7189")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Based on the function signature and implementation, we need:
# - tls_client_cert: a certificate value (string or file path)
# - tls_client_key: a key value (string or file path)
# - tls_client_passwd: a password value (string or None)

# The function checks if tls_client_cert is truthy and returns a tuple if so
# We'll use realistic SSL/TLS certificate and key representations

ARGS = {
    "tls_client_cert": "/path/to/client.crt",
    "tls_client_key": "/path/to/client.key",
    "tls_client_passwd": "password123"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
macos_requirement Ran
ran on a synthesized input
instructlab/instructlab/src/instructlab/utils.py
code served (permissive licence) · get_code("6f0eee9a6bc07bb6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned function. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn
from functools import wraps


class MockEchoFunc:
    """Mock echo function that accepts message and fg parameters"""
    def __call__(self, message, fg='white'):
        pass


class MockExitException(Exception):
    """Mock exit exception"""
    def __init__(self, code):
        self.code = code
        super().__init__(f"Exit with code {code}")


def echo_func(message, fg='white'):
    """Simple echo function for testing"""
    pass


def exit_exception(code):
    """Simple exit exception for testing"""
    raise MockExitException(code)


ARGS = {
    'echo_func': echo_func,
    'exit_exception': MockExitException
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
make_lab_diff_aliases Not yet run
unverified (harvested, no recorded run)
instructlab/instructlab/src/instructlab/utils.py
code served (permissive licence) · get_code("9ec8434d39e817bf")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

This work introduces LAB (Large-scale Alignment for chatBots), a novel methodology designed to overcome the scalability challenges in the instruction-tuning phase of large language model (LLM) training. Leveraging a taxonomy-guided synthetic data generation process and a multi-phase tuning framework, LAB significantly reduces reliance on expensive human annotations and proprietary models like GPT-4. We demonstrate that LAB-trained models can achieve competitive performance across several benchmarks compared to models trained with traditional human-annotated or GPT-4 generated synthetic data. Thus offering a scalable, cost-effective solution for enhancing LLM capabilities and instruction-following behaviors without the drawbacks of catastrophic forgetting, marking a step forward in the efficient training of LLMs for a wide range of applications.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2403.01081")
get_code_for_paper("2403.01081")
have("2403.01081")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2403.01081.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2403.01081.svg)](https://syntology.ai/paper/2403.01081)

Connect an agent — have() is free.